Perception

It is argued that while pictorial cues must be fundamental in the perception of an extended ground plane in daily life (and, by definition, in the perception of depth in pictures), the traditional pictorial cues such as perspective or texture gradients are neither a necessary nor a sufficient basis for it. That they are not necessary was shown by experiments in which such cues were eliminated from pictures representing a scene in depth. Illusory size perception based on localization of objects in depth in such pictures nonetheless occurred. That they are not sufficient was shown by experiments in which photographs of grassy fields did not yield impressions of depth or related size illusions when conditions were such that the scene was not recognized. Once recognized, however, these same pictures did yield such perceptions. It is suggested that a critical step in perceiving depth based on pictorial information is recognition of the scene.

An attempt is made to list the visual phenomena exploited in op art. These include moiré fringes, afterimages, Hermann grid effects, Gestalt grouping principles, blurring and movement due to astigmatic fluctuations in accommodation, scintillation and streaming possibly due to eye movements, and visual persistence. The historical origins of these phenomena are also noted.

The role of apparent depth features and the proximity of the test lines to the adjacent contours in the actuation of the Ponzo illusion was examined. Six versions of the Ponzo figure were employed: a standard Ponzo figure and five modified figures in which the test lines varied in orientation (horizontal or vertical) and in location (inside or outside the converging contours). Both manipulations resulted in a significant decrease in the magnitude of the illusion in comparison to the standard Ponzo figure. The results suggest that the Ponzo illusion is significantly affected by contextual factors.

Five experiments which attempted to evaluate the relationship between orientation and curvature selectivity in human vision are described. In the first two experiments, threshold elevation for curved gratings was measured after exposure to similar gratings, with the use of either an adaptation (experiment 1) or a masking (experiment 2) paradigm. In both experiments threshold elevation occurred which was selective for both the degree and the direction of curvature of the adapting pattern. Experiment 3 compared the effects of adapting to tilted rectilinear or vertical curved gratings upon threshold for a vertical rectilinear grating. Threshold elevation declined systematically as the adapting gratings were either tilted or made more curved. Experiment 4 measured curvature selectivity as a function of the orientation of a curved adapting grating. Threshold elevation declined as the adapting grating was tilted more, but curvature selectivity remained. Experiment 5 measured the orientation tuning for curved gratings directly. Threshold elevation declined to 50% of its maximum value at an adapting orientation of about 28°. This was constant for all values of curvature used. The results are discussed with reference to the question of whether the human visual system contains ‘curvature detectors’ or linear-contour detectors which respond to the tangents of curves.

The decay of several visual aftereffects may be prolonged by interposing a period of light-free or pattern-free viewing between adaptation and testing. We demonstrate that this storage phenomenon can be observed using the threshold elevation aftereffect that follows inspection of a high-contrast grating pattern. Control experiments comparing thresholds for vertical and horizontal gratings after adaptation to a vertical grating reveal that the stored aftereffect, like its unstored counterpart, is pattern-selective. Storage is equally pronounced with stimuli that are detected by pattern-analyzing or movement-analyzing visual channels. Unlike other aftereffects, the threshold-elevation aftereffect requires that the storage period be light-free; no storage is seen if a blank field is inspected between adaptation and testing. The results are discussed with respect to the nature of visual aftereffects, and possible cognitive or physiological models of storage.

Three experiments were conducted to test the hypothesis that the apparent movement of a stimulus when the two eyes, are alternately occluded can be explained by phoria and Hering's principles of visual direction. In experiments 1 and 2, the direction of apparent movement, eye position, and eye movements were measured when eye movements did and did not occur. In experiment 3, the magnitude of apparent movement and the extent of phoria were compared. Results from experiments 1 and 2 indicated that the direction of apparent movement could be predicted from the direction of phoria, in conjunction with Hering's principles, and was not contingent on eye movements. In experiment 3, a high positive correlation (r = 0.95) between the magnitude of apparent movement and the extent of phoria was obtained.

We report a striking visual illusion which involves an induced motion that is delayed in time. An observer visually tracks a moving target in the neighborhood of a fixed target. The fixed target appears to be entrained by the moving target and appears to follow its movements after a lag of 0.33 second. A series of experiments showed that while the illusion depends on low background salience it obtains with both smooth and oscillatory motion in all directions including depth, under monocular and binocular viewing, in the absence of vestibular and kinesthetic motion cues or eye movements, and under a wide range of relative and absolute target luminance and position. The strength of the illusion and the magnitude of the induced movement's delay seem not to depend on any of the above factors. The illusion, by resisting a peripheral explanation, may provide some clues to central perceptual processing. In addition, we suggest the possibility that errors of judgment based on the nonveridical perception of motion displaced in time may play a role in nighttime automobile and aircraft accidents.

The accuracy of many perceptual comparisons depends greatly on the order in which the to-be-compared stimuli are presented. With comparisons of durations around 300 ms, these presentation-order effects do not diminish, even with extended practice, when feedback about response accuracy is withheld. Providing such feedback greatly diminishes presentation-order effects and coincidentally produces substantial increases in response accuracy. The feedback acts in part through inducing response biases and in part through changes in sensitivity. The contradiction between studies which report time-order errors in duration comparison and those which do not is attributable to differences in the use of information feedback.

Pandemonium-like models have played a central role in theories of perceptual recognition. One model is examined which asserts that information is sorted unidirectionally through a hierarchy of increasingly abstract levels only to a depth required by the logical demands of the task and read off from the appropriate level to control response decisions. The support originally claimed for the model in terms of its application to visual search performance is questioned. It is suggested that the pervasiveness of such models is not due to their competition with alternative theories but rather to methatheoretic considerations.

A series of seven photographs which depict a real scene as it is approached by an observer were shown, in sequence, to ninety-six university students. The photographs were either projected colour transparencies or black-and-white prints. Two sizes of photographs were used in each set and the series was presented with or without the first photograph. In earlier studies in which blurred photographs were used as stimuli, observers were unsure about their hypotheses. That was not the case in this study. Although individual observers differed widely in their initial hypotheses, the initial photographs were not perceived as ambiguous. It required on average from 2.9 photographs (large, black-and-white, full series) to 4.5 photographs (small, colour, full series) for correct identification. In general, colour is a hindrance to correct identification. For the full series approximately three hypotheses are put forward before the scene is correctly identified.

Several numerical and graphical statistical methods are illustrated in an analysis of data from an experiment that investigated a hypothesis of Julesz that giving a person a priori information about the structure of a complex random-dot stereogram reduces the time needed to perceive it when it is viewed. The data are divided into two groups, one consisting of those observers who received no cue or verbal cues (NV) and the other consisting of those who received verbal-visual cues (VV). A quantile-quantile plot shows that the NV times (mean = 7.6) are longer than the VV times (mean = 5.6). By using probability plots, it is shown that the perception times have an exponential probability distribution. A hypothesis test based upon this distribution is used to show that the differences between the NV and W times has significance slightly below 0.05.

Patients whose vision has been restored after long periods of blindness commonly experience difficulty in perceiving. Parodoxically, such patients often perform well on pseudoisochromatic colour-vision tests. A well-known example is Gregory and Wallace's subject, S B, whose performance on the Ishihara test was perfect. It is suggested that restoration of vision may be associated with relatively poor visual acuity and that this, by filtering higher spatial frequencies, enhances the legibility of pseudoisochromatic test patterns. Two experiments confirm that the Ishihara plates are more legible when seen as defocused images.